On Thursday, November 26, 2020 at 1:28:53 PM UTC+1 [email protected] wrote:
> The architecture they specify in their paper > > https://arxiv.org/pdf/2011.09850.pdf > > could be of interest to the AI hacker (MIT AI Lab type) trying to > implement and demonstrate a "conscious" robot" (even just an impressively > "fake" one). > > *We agree with Christof Koch that “There isn’t a Turing Test for > consciousness. You have to look at the way the system is built. You have to > look at the circuitry, not [only] its behavior”. We would emphasize > “architecture” over “circuitry”.* > Show me a set of systems that demonstrate indeterminacy between pursuing some choice of goal(s) and doing nothing. If they demonstrate activity wherein a goal is repurposed or glitched as an end in itself; i.e. where the goal's completion/felicity becomes irrelevant to the continued pursuit of the activity, which itself demonstrates constant refinement and increased specificity at each iteration as in social interaction, sex, drugs, rock and roll etc. you're closer to a conscious machine, regardless of architecture, circuitry, way the system is built etc. Particularly if it asks the admin for more resources to pursue its ends. One difficulty lies in demonstrating that the system(s) achieved such freely; hence you have to show that some significant number of them did nothing and/or other things. PGC -- You received this message because you are subscribed to the Google Groups "Everything List" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/everything-list/a62c0f1a-9c70-4a24-978f-10ac3bb9daf7n%40googlegroups.com.
